Various chemotherapeutic regimens are often employed in the treatment of human neoplastic diseases. As exemplified by the frequent administration of vincristine, chemical agents demonstrating a high degree of clinical efficacy occur naturally in plants. Characterization of the molecular structures of (or biochemical mechanisms facilitated by) naturally occuring substances has also largely contributed to the clinical success demonstrated by antineoplastic agents produced synthetically. We currently propose to isolate and structurally characterize endogenous antineoplastic agents from approximately 24 plants. Individual crude extracts derived from these plants are known to enhance the survival time of P388 leukemic mice, suggesting that the active principle(s) may be of practical use for the treatment of cancer. On the basis of chemotaxonomic relationships and the global regions from which these plants were derived, it can be predicted that the pure, biologically active substances should be structurally unique, and perhaps function by novel biochemical mechanisms. Routine methods of isolation and structure elucidation will generally be employed. Additionally, however, innovative procedures of isolation will be investigated which may represent significant contributions to the field. For monitoring the effectiveness of the purification procedures and quantitating the biological activity of the resulting pure isolates, in vitro mammalian cell culture assays will be used. Attempts will be made to enhance the correlation of results obtained with in vitro cell culture assays with those obtained with in vivo antitumor assays by utilization of enzyme systems capable of metabolizing (potential) antitumor agents to active species. Conversely, metabolic inactivation may be observed and this should be equally relevant to the predication of results which would be obtained following administration of the agent to intact animals. The metabolites of pure compounds demonstrating altered biological activity when subjected to these experimental conditions will be structurally characterized to provide information relevant to structure-activity relationships which may indicate that synthetic derivatives could be prepared that would be highly active antitumor agents. Finally, suitable quantities of sufficiently active antineoplastic agents will be supplied for evaluation in systems such as the National Cancer Institute tumor panel; this represents an initial step toward the ultimate goal of providing a chemotherapeutic agent worthy of human administration.